The ANR, by supporting the PreSPIN project, recognises the excellence and potential of the research carried out by our university.

PreSPIN: Predictive simulation for the planning of interventional neuroradiology procedures

The interactive simulation of soft objects and fluids now enables commercial endovascular simulators to offer a high degree of realism suitable for the training of practitioners. To further integrate simulation into the therapeutic workflow, there is a need to design patient-specific simulation models and algorithms that go far beyond mere realism and achieve a high level of numerical precision. However, compared to experimental and in vivo measurements, advanced algorithms are still far from being able to capture the complexity and subtleties of real phenomena. The PreSPIN project aims to bridge the gap between training and intervention by tackling the planning phase. Here, digital simulation can play a fundamental role in choosing the best treatment strategy, optimising the use of medical resources and better preparing medical teams for potential risks, with reduced intervention time and better patient outcomes. However, this is provided that the simulation is predictive, that is to say capable of reproducing the surgical events that are essential for doctors. Because of this dependency on the application, PreSPIN will focus on the endovascular treatment of ischemic strokes. In this context, PreSPIN will develop new solutions for the modelling of the cerebrovascular system from MRA (magnetic resonance angiography) data, which will provide geometric boundary conditions for future simulations; improve interactive navigation of intervention devices, for a more informed choice of medical devices and treatment options; move towards rapid simulation of blood flow, to prevent potential risks associated with restoration of blood flow; and provide new and improved means of simulating perfusion imaging to better predict the treatment outcome. These advances will be supported by acquisition of experimental data with the aim of providing the community with in vitro data and open-source reproducible configurations. This data will be developed in close collaboration with interventional neuroradiologists, in order to ensure that critical phenomena are captured and to progress towards a technical definition and evaluation of medical predictability.

The PreSPIN project is led by Inria GE (PI: Erwan Kerrien), and is based on a consortium of four partners (Inria GE, URCA, Insa Lyon, Nancy University Hospital). For URCA (director: N. Passat), the units involved are the CReSTIC and the LMR, which rely on significant experience acquired during previous ANR projects carried out by these two research units: VIVABRAIN (2012, PI: N. Passat), MAIA (2015), Hanuman (2018: PI : S. Salmon) and R-Vessel-X (2019).